写在前面的话

如果我有一台基于云的虚拟机，并且运行了Ubuntu或CentOS的Linux服务器系统，而这些设备可能是虚拟化的，也可能不是虚拟化的，那我们应该如何去判断这台Linux Guest系统设备使用的是哪种虚拟化技术（VMWARE/ KVM/ XEN/ VirtualBox/ Container/lxc/Hyper-V等等）呢？那我们如何才能识别一台CentOS 7 Linux VPS所使用的虚拟化技术类型呢？

技术分析

我们需要使用virt-what程序（一个Shell脚本）来检测目标设备系统所使用的虚拟化技术类型，它可以打印出每一条关于目标设备虚拟化技术的相关信息。在这篇文章中，我们将介绍如何去判断Linux Guest VM虚拟化技术类型。

判断Linux Guest VM虚拟化技术类型

正如我们刚才说的，我们需要使用virt-what Linux命令来判断一个程序当前是否在一台虚拟机设备上运行。该脚本支持各种虚拟机管理程序，接下来，我们看看如何在不同的Linux发行版系统中安装virt-what。

Debian或Ubuntu Linux安装

首先，运行下列apt命令/apt-get命令：

$ sudo apt-get install virt-what

或者

$ sudo apt install virt-what

在RHEL/CentOS/Scientific Linux VM中安装virt-what

我们可以使用yum命令完成安装：

$ sudo yum install virt-what

Fedora Linux VM安装virt-what来检测当前环境是否为虚拟机环境

执行dnf命令：

$ sudo dnf install virt-what

Suse/OpenSUSE安装

运行zypper命令：

sudo zypper in virt-what

判断远程服务器是否为虚拟机环境

我们可以直接输入并运行下列命令：

$ $ sudo virt-what
    xen
    xen-domU

输出结果：

$ $ sudo virt-what
    kvm

其他输出结果：

$ $ sudo virt-what
    lxc

下面给出的是AWS返回的结果：

$ sudo virt-what
    xen
    xen-hvm
    aws

其他可能的值

1、hyperv : 这是微软的Hyper-V管理程序；

2、parallels : 访客用户正在Parallels虚拟平台（Parallels Desktop、Parallels Server）中运行；

3、powervm_lx86 : 访客用户正在IBM powervm lx86 Linux/x86模拟器中运行；

4、qemu : 这是使用软件模拟技术的QEMU管理程序；

5、virtualpc : 访客用户正在微软VirualPC上运行；

6、xen-hvm : 这是一个Xen Guest完全虚拟化平台（HVM）；

7、uml : 这是一个用户模式Linux（UML）客户机；

8、openvz : 访客用户正在OpenVZ或Virtuozzo容器中运行；

9、linux_vserver : 此进程正在Linux VServer容器中运行；

10、ibm_systemz : 这是一个IBM SystemZ硬件分区系统；

如果没有任何输出结果，那意味着我们的程序可能是在裸机上运行的，或者程序是在一种我们无法检测到的虚拟机系统中运行的。

检测脚本源代码

下面给出的是脚本的源代码：

#!/bin/sh -
    # virt-what.  Generated from virt-what.in by configure.
    # Copyright (C) 2008-2017 Red Hat Inc.
    #
    # This program is free software; you can redistribute it and/or modify
    # it under the terms of the GNU General Public License as published by
    # the Free Software Foundation; either version 2 of the License, or
    # (at your option) any later version.
    #
    # This program is distributed in the hope that it will be useful,
    # but WITHOUT ANY WARRANTY; without even the implied warranty of
    # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    # GNU General Public License for more details.
    #
    # You should have received a copy of the GNU General Public License
    # along with this program; if not, write to the Free Software
    # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
    # 'virt-what' tries to detect the type of virtualization being
    # used (or none at all if we're running on bare-metal).  It prints
    # out one or more lines each being a 'fact' about the virtualization.
    #
    # Please see also the manual page virt-what(1).
    # This script should be run as root.
    #
    # The following resources were useful in writing this script:
    # . http://dmo.ca/blog/detecting-virtualization-on-linux/
    # Do not allow unset variables, and set defaults.
    set -u
    root=''
    skip_qemu_kvm=false
    skip_lkvm=false
    VERSION="1.20"
    have_cpuinfo () {
        test -e "${root}/proc/cpuinfo"
    }
    use_sysctl() {
        # Lacking /proc, on some systems sysctl can be used instead.
        OS=$(uname) || fail "failed to get operating system name"
        [ "$OS" = "OpenBSD" ]
    }
    fail () {
        echo "virt-what: $1" >&2
        exit 1
    }
    usage () {
        echo "virt-what [options]"
        echo "Options:"
        echo "  --help          Display this help"
        echo "  --version       Display version and exit"
        exit 0
    }
    # Handle the command line arguments, if any.
    while test $# -gt 0; do
        case "$1" in
            --help) usage ;;
            --test-root=*)
                # Deliberately undocumented: used for 'make check'.
                root=$(echo "$1" | sed 's/.*=//')
                shift 1
                test -z "$root" && fail "--test-root option requires a value"
                ;;
            -v|--version) echo "$VERSION"; exit 0 ;;
            --) shift; break ;;
            *) fail "unrecognized option '$1'";;
        esac
    done
    test $# -gt 0 && fail "extra operand '$1'"
    # Add /sbin and /usr/sbin to the path so we can find system
    # binaries like dmidecode.
    # Add /usr/libexec to the path so we can find the helper binary.
    prefix=/usr/local
    exec_prefix=${prefix}
    PATH="${root}${exec_prefix}/libexec:${root}/sbin:${root}/usr/sbin:${PATH}"
    export PATH
    # Check we're running as root.
    EFFUID=$(id -u) || fail "failed to get current user id"
    if [ "x$root" = "x" ] && [ "$EFFUID" -ne 0 ]; then
        fail "this script must be run as root"
    fi
    # Try to locate the CPU-ID helper program
    CPUID_HELPER=$(which virt-what-cpuid-helper 2>/dev/null)
    if [ -z "$CPUID_HELPER" ] ; then
        fail "virt-what-cpuid-helper program not found in \$PATH"
    fi
    # Many fullvirt hypervisors give an indication through CPUID.  Use the
    # helper program to get this information.
    cpuid=$(virt-what-cpuid-helper)
    # Check for various products in the BIOS information.
    # Note that dmidecode doesn't exist on all architectures.  On the ones
    # it does not, then this will return an error, which is ignored (error
    # message redirected into the $dmi variable).
    dmi=$(LANG=C dmidecode 2>&1)
    # Architecture.
    # Note for the purpose of testing, we only call uname with -m option.
    arch=$(uname -m | sed -e 's/i.86/i386/' | sed -e 's/arm.*/arm/')
    # Check for VMware.
    # cpuid check added by Chetan Loke.
    if [ "$cpuid" = "VMwareVMware" ]; then
        echo vmware
    elif echo "$dmi" | grep -q 'Manufacturer: VMware'; then
        echo vmware
    fi
    # Check for Hyper-V.
    # http://blogs.msdn.com/b/sqlosteam/archive/2010/10/30/is-this-real-the-metaphysics-of-hardware-virtualization.aspx
    if [ "$cpuid" = "Microsoft Hv" ]; then
        echo hyperv
    fi
    # Check for VirtualPC.
    # The negative check for cpuid is to distinguish this from Hyper-V
    # which also has the same manufacturer string in the SM-BIOS data.
    if [ "$cpuid" != "Microsoft Hv" ] &&
        echo "$dmi" | grep -q 'Manufacturer: Microsoft Corporation'; then
        echo virtualpc
    fi
    # Check for VirtualBox.
    # Added by Laurent Léonard.
    if echo "$dmi" | grep -q 'Manufacturer: innotek GmbH'; then
        echo virtualbox
    fi
    # Check for bhyve.
    if [ "$cpuid" = "bhyve bhyve " ]; then
      echo bhyve
    elif echo "$dmi" | grep -q "Vendor: BHYVE"; then
      echo bhyve
    fi
    # Check for OpenVZ / Virtuozzo.
    # Added by Evgeniy Sokolov.
    # /proc/vz - always exists if OpenVZ kernel is running (inside and outside
    # container)
    # /proc/bc - exists on node, but not inside container.
    if [ -d "${root}/proc/vz" -a ! -d "${root}/proc/bc" ]; then
        echo openvz
    fi
    # Check for LXC containers
    # http://www.freedesktop.org/wiki/Software/systemd/ContainerInterface
    # Added by Marc Fournier
    if [ -e "${root}/proc/1/environ" ] &&
        cat "${root}/proc/1/environ" | tr '\000' '\n' | grep -Eiq '^container='; then
        echo lxc
    fi
    # Check for Linux-VServer
    if test -e "${root}/proc/self/status" \
       && cat "${root}/proc/self/status" | grep -q "VxID: [0-9]*"; then
        echo linux_vserver
        if grep -q "VxID: 0$" "${root}/proc/self/status"; then
            echo linux_vserver-host
        else
            echo linux_vserver-guest
        fi
    fi
    # Check for UML.
    # Added by Laurent Léonard.
    if have_cpuinfo && grep -q 'UML' "${root}/proc/cpuinfo"; then
        echo uml
    fi
    # Check for IBM PowerVM Lx86 Linux/x86 emulator.
    if have_cpuinfo && grep -q '^vendor_id.*PowerVM Lx86' "${root}/proc/cpuinfo"
    then
        echo powervm_lx86
    fi
    # Check for Hitachi Virtualization Manager (HVM) Virtage logical partitioning.
    if echo "$dmi" | grep -q 'Manufacturer.*HITACHI' &&
       echo "$dmi" | grep -q 'Product.* LPAR'; then
        echo virtage
    fi
    # Check for IBM SystemZ.
    if have_cpuinfo && grep -q '^vendor_id.*IBM/S390' "${root}/proc/cpuinfo"; then
        echo ibm_systemz
        if [ -f "${root}/proc/sysinfo" ]; then
            if grep -q 'VM.*Control Program.*KVM/Linux' "${root}/proc/sysinfo"; then
                echo ibm_systemz-kvm
            elif grep -q 'VM.*Control Program.*z/VM' "${root}/proc/sysinfo"; then
                echo ibm_systemz-zvm
            elif grep -q '^LPAR' "${root}/proc/sysinfo"; then
                echo ibm_systemz-lpar
            else
                # This is unlikely to be correct.
                echo ibm_systemz-direct
            fi
        fi
    fi
    # Check for Parallels.
    if echo "$dmi" | grep -q 'Vendor: Parallels'; then
        echo parallels
        skip_qemu_kvm=true
    fi
    # Check for oVirt/RHEV.
    if echo "$dmi" | grep -q 'Manufacturer: oVirt'; then
        echo ovirt
    fi
    if echo "$dmi" | grep -q 'Product Name: RHEV Hypervisor'; then
        echo rhev
    fi
    # Check for Xen.
    if [ "$cpuid" = "XenVMMXenVMM" ] &&
        ! echo "$dmi" | grep -q 'No SMBIOS nor DMI entry point found, sorry'; then
        echo xen; echo xen-hvm
        skip_qemu_kvm=true
    elif [ -d "${root}/proc/xen" ]; then
        echo xen
        if grep -q "control_d" "${root}/proc/xen/capabilities" 2>/dev/null; then
            echo xen-dom0
        else
            echo xen-domU
        fi
        skip_qemu_kvm=true
        skip_lkvm=true
    elif [ -f "${root}/sys/hypervisor/type" ] &&
        grep -q "xen" "${root}/sys/hypervisor/type"; then
        # Ordinary kernel with pv_ops.  There does not seem to be
        # enough information at present to tell whether this is dom0
        # or domU.  XXX
        echo xen
    elif [ "$arch" = "arm" ] || [ "$arch" = "aarch64" ]; then
        if [ -d "${root}/proc/device-tree/hypervisor" ] &&
            grep -q "xen" "${root}/proc/device-tree/hypervisor/compatible"; then
            echo xen
            skip_qemu_kvm=true
            skip_lkvm=true
        fi
    elif [ "$arch" = "ia64" ]; then
        if [ -d "${root}/sys/bus/xen" -a ! -d "${root}/sys/bus/xen-backend" ]; then
            # PV-on-HVM drivers installed in a Xen guest.
            echo xen
            echo xen-hvm
        else
            # There is no virt leaf on IA64 HVM.  This is a last-ditch
            # attempt to detect something is virtualized by using a
            # timing attack.
            virt-what-ia64-xen-rdtsc-test > /dev/null 2>&1
            case "$?" in
                0) ;; # not virtual
                1) # Could be some sort of virt, or could just be a bit slow.
                    echo virt
            esac
        fi
    fi
    # Check for QEMU/KVM.
    #
    # Parallels exports KVMKVMKVM leaf, so skip this test if we've already
    # seen that it's Parallels.  Xen uses QEMU as the device model, so
    # skip this test if we know it is Xen.
    if ! "$skip_qemu_kvm"; then
        if [ "$cpuid" = "KVMKVMKVM" ]; then
            echo kvm
        elif [ "$cpuid" = "TCGTCGTCGTCG" ]; then
            echo qemu
            skip_lkvm=true
        elif echo "$dmi" | grep -q 'Product Name: KVM'; then
            echo kvm
            skip_lkvm=true
        elif echo "$dmi" | grep -q 'Manufacturer: QEMU'; then
            # The test for KVM above failed, so now we know we're
            # not using KVM acceleration.
            echo qemu
            skip_lkvm=true
        elif [ "$arch" = "arm" ] || [ "$arch" = "aarch64" ]; then
            if [ -d "${root}/proc/device-tree" ] &&
                ls "${root}/proc/device-tree" | grep -q "fw-cfg"; then
                # We don't have enough information to determine if we're
                # using KVM acceleration or not.
                echo qemu
                skip_lkvm=true
            fi
        elif [ -d ${root}/proc/device-tree/hypervisor ] &&
             grep -q "linux,kvm" /proc/device-tree/hypervisor/compatible; then
            # We are running as a spapr KVM guest on ppc64
            echo kvm
            skip_lkvm=true
        elif use_sysctl; then
            # SmartOS KVM
            product=$(sysctl -n hw.product)
            if echo "$product" | grep -q 'SmartDC HVM'; then
                echo kvm
            fi
        else
            # This is known to fail for qemu with the explicit -cpu
            # option, since /proc/cpuinfo will not contain the QEMU
            # string. QEMU 2.10 added a new CPUID leaf, so this
            # problem only triggered for older QEMU
            if have_cpuinfo && grep -q 'QEMU' "${root}/proc/cpuinfo"; then
                echo qemu
            fi
        fi
    fi
    if ! "$skip_lkvm"; then
        if [ "$cpuid" = "LKVMLKVMLKVM" ]; then
            echo lkvm
        elif [ "$arch" = "arm" ] || [ "$arch" = "aarch64" ]; then
            if [ -d "${root}/proc/device-tree" ] &&
                grep -q "dummy-virt" "${root}/proc/device-tree/compatible"; then
                echo lkvm
            fi
        fi
    fi
    # Check for Docker.
    if [ -f "${root}/.dockerinit" ]; then
        echo docker
    fi
    # Check ppc64 lpar, kvm or powerkvm
    # example /proc/cpuinfo line indicating 'not baremetal'
    # platform  : pSeries
    #
    # example /proc/ppc64/lparcfg systemtype line
    # system_type=IBM pSeries (emulated by qemu)
    if [ "$arch" = "ppc64" ] || [ "$arch" = "ppc64le" ] ; then
        if have_cpuinfo && grep -q 'platform.**pSeries' "${root}/proc/cpuinfo"; then
            if grep -q 'model.*emulated by qemu' "${root}/proc/cpuinfo"; then
                    echo ibm_power-kvm
            else
                # Assume LPAR, now detect shared or dedicated
                if grep -q 'shared_processor_mode=1' "${root}/proc/ppc64/lparcfg"; then
                    echo ibm_power-lpar_shared
                else
                    echo ibm_power-lpar_dedicated
                fi
            # detect powerkvm?
            fi
        fi
    fi
    # Check for OpenBSD/VMM
    if [ "$cpuid" = "OpenBSDVMM58" ]; then
            echo vmm
    fi
    # Check for LDoms
    if [ "${arch#sparc}" != "$arch" ] && [ -e "${root}/dev/mdesc" ]; then
        echo ldoms
        if [ -d "${root}/sys/class/vlds/ctrl" ] && \
                 [ -d "${root}/sys/class/vlds/sp" ]; then
            echo ldoms-control
        else
            echo ldoms-guest
        fi
        MDPROP="${root}/usr/lib/ldoms/mdprop.py"
        if [ -x "${MDPROP}" ]; then
            if [ -n "$($MDPROP -v iodevice device-type=pciex)" ]; then
                echo ldoms-root
                echo ldoms-io
            elif [ -n "$($MDPROP -v iov-device vf-id=0)" ]; then
                echo ldoms-io
            fi
        fi
    fi
    # Check for AWS.
    # AWS on Xen.
    if echo "$dmi" | grep -q 'Version: [0-9]\.[0-9]\.amazon'; then
        echo aws
    # AWS on baremetal or KVM.
    elif echo "$dmi" | grep -q 'Vendor: Amazon EC2'; then
        echo aws
    fi

如何使用dmidecode命令来寻找相同的信息

配合Bash for循环并使用dmidecode命令判断目标虚拟化技术：

for i in system-manufacturer system-product-name
    do 
      sudo dmidecode -s $i
    done

样本输出：

Red Hat
    KVM

其他用来判断虚拟化技术的命令

我们还可以使用systemd-detect-virt命令来检测虚拟化环境：

systemd-detect-virt

资源获取

1、virt-what：【点我获取】

2、Bash for循环：【参考文档】

3、yum命令：【参考文档】

4、apt命令：【参考文档】

5、apt-get命令：【参考文档】

* 参考来源：cyberciti，FB小编Alpha_h4ck编译，转载请注明来自FreeBuf.COM